fixed point tool collect ranges problem

The input of the subsystem to be converted to fixed point in double and specified data types differ significantly,almost one thousand times.And there are many overflows in double mode,and few overflows in specified data types mode.Looking forward to your reply,thank you.

[Walter Roberson]

Overflows in double mode? Do you mean that the values exceed

fprintf('%lu\n', flintmax())

9007199254740992

or do you mean that the values exceed

fprintf('%.16g\n', realmax())

1.797693134862316e+308

??

What kind of range of data are you using?

[Z]

I found there is some data type convert before the subsystem and the data type convert overflow. I made mistake when I built the model.